`

Postsecondary

Non-Completion Among Veterans: Contributing Factors and Implications

November 2018

2

ABOUT VETERANS EDUCATION SUCCESS

Veterans Education Success (VES) is a nonprofit organization dedicated to protecting and defending the

integrity and promise of the GI Bill and other federal education programs for veterans and

servicemembers.1 VES provides:

• Research: Non-partisan research on issues of concern to student veterans, including outcomes and

student loan debt levels.

• Free Help: Free legal services, advice, and college and career counseling for veterans,

servicemembers, survivors, and their families who faced fraud or abuse from postsecondary

institutions when using their GI Bill benefits.

• Civic Engagement: Help for veterans in exercising their democratic rights by engaging their

Congressional representatives, federal agencies, and local media, including speaking out at public

hearings.

• Policy and Advocacy: Assistance to federal policymakers to improve quality in higher education and

protect the integrity of the GI Bill and other student aid.

• Whistleblowers: Free assistance to college whistleblowers in exposing fraud, and free assistance to

federal and state law enforcement agencies in helping to stop consumer fraud by postsecondary

institutions.

1GI Bill® is a registered trademark of the U.S. Department of Veterans Affairs. More information about education benefits offered by VA is available at the official U.S. government Web site at https://www.benefits.va.gov/gibill.

3

CONTENTS Executive Summary Background Findings

Data May Overstate Veteran Non-Completion Rates

Demographics of Veterans Who Left Factors Contributing to Non-Completion Implications of Non-Completion

Conclusions References Data Sources and Methodology Note. All data in this report are based on the authors’ analyses of the Department of Education’s longitudinal Beginning Postsecondary Students (BPS), which surveyed first-time students in 2011-12 and again in June 2014. We examined non-completion among student veterans and a comparison group of non-veteran, independent students. Throughout this paper, we refer to non-veteran, independent students as non-veterans. Sample sizes were generally too small to examine outcomes separately for certificates vs. associate’s degrees and for students who attended non-profit institutions. After only 3-years of attendance, most students seeking bachelor’s degrees were still enrolled. Acknowledgements. The authors wish to thank Dr. Dani Molina who reviewed a draft of this report and provided thoughtful feedback. Dr. Molina is a uniquely qualified reviewer. He is a veteran who began his post-secondary studies after he was discharged, earning a bachelor’s degree from the University of California, Santa Cruz and a master’s and PhD from the University of California, Los Angeles. He served as senior program and research manager for veterans’ programs at the American Council on Education and is currently the Director of the Veterans Resource Center at California State University, Los Angeles, where he focuses on services that help veterans make a successful transition from service to school.

4

EXECUTIVE SUMMARY All data in this report are based on the authors’ analyses of the Department of Education’s longitudinal Beginning Postsecondary Students (BPS), which surveyed first-time students twice from 2011 to 2014. BPS May Overstate Veteran Non-Completion. By most measures, student veterans demonstrated stronger rates of persistence and completion than non-veterans.

• Twenty percent of veterans who first enrolled in postsecondary education in 2011-12 had left without a degree by 2014 compared to 40 percent of non-veterans.

• Twenty-eight percent of veterans had earned a certificate or associate’s degree compared to 23 percent of their non-veteran counterparts. Moreover, about half were still enrolled 3 years after starting school while only about one third of non-veterans were still pursuing a degree.

BPS may overstate veteran non-completion because of the somewhat greater propensity of student veterans to take breaks and then reenroll, a circumstance that makes it difficult to say with certainty that they will not at some point return to earn a degree. This uncertainty is underscored by the fact that 40 percent of veterans who completed certificates or degrees in 2015-16 began their postsecondary education in 2005 or earlier compared to about 30 percent of non-veterans.2 Demographics of Non-Completers Align with Risk Factors. The demographics of veterans who left without a degree by 2014 are closely associated with risk factors often linked to non-completion. Veterans who left higher education:

• were more likely to be first in their family to go to college, lack a traditional high-school diploma, be disabled, be single or married with dependents, and work full-time.3

• exhibited 5 or more risk factors associated with non-completion compared to their peers who earned certificates or degrees—21 percent vs. 4 percent, respectively.

A greater proportion of both non-veterans who left higher education and non-veterans who graduated had 5 or more risk factors, but the proportion for non-veterans who graduated was more than 4 times larger than for veterans who earned a certificate or degree—17 percent vs. 4 percent.

2Because BPS graduation data is limited to those who earned a degree in just 3-years, we used another Education Department Survey, the National Postsecondary Student Aid Study.

3When evaluating risk factors that may adversely affect persistence and attainment for nontraditional students, BPS classifies individuals with a GED as having “no high school diploma.”

5

Although not a recognized risk factor, research suggests that online courses are the most difficult for those who are the least prepared leading to worse outcomes than brick and mortar classes. Because most veteran undergraduates use their GI Bill benefits after discharge from active duty, they may not have been in a classroom setting since leaving high school and may require remedial coursework.4 Moreover, online classes are appealing to older, adult learners who are juggling family and work responsibilities.5 The proportion of veterans and non-veterans who left after taking classes exclusively online was similar but the divergence between non-completion at for-profit and public institutions was stark.

• Thirty six percent of veterans enrolled exclusively online at for-profit schools had left by 2014 compared to 19 percent at public institutions.

• Non-veterans exhibited a similar intra-sector pattern, but much smaller proportions left—22 percent of those attending for-profit exclusively online programs vs. 9 percent of those enrolled exclusively online at public institutions.

Implications of Non-Completion. Although research suggests that non-graduates are more likely than graduates to face repayment challenges, it also suggests that student loan debt need not be excessive to cause financial stress in the near term for recent graduates and in the longer term for those who leave without a degree. In other words, even small amounts of debt can be a barrier to repayment among college graduates and non-completers. Compared to veterans, larger proportions of non-veterans incurred federal or private student loan debt, including those who left without a degree.

• Seventy-two percent of veterans who left had no student loan debt, possibly a result of the new, more generous Post-9/11 GI Bill and a reflection of the considerable proportion of veterans who attend inexpensive community colleges. In contrast, 54 percent of non-veterans who left without a degree had no student loan debt, an 18-percentage point difference.

• A larger proportion of veterans who left for-profit schools without a degree had student loan debt compared to their peers at public institutions—36 percent vs. 24 percent, respectively.

Despite monthly payments of $100 or less for 65 percent of veterans and 73 percent of non-veterans who left without a degree, about three quarters of each have had a forbearance and almost all have had at least one delinquency. Moreover, 19 percent of veterans and 12 percent of non-veterans had been delinquent 6 or more times, suggesting that a larger share of student veterans might be at risk of defaulting. Repayment challenges can begin prior to graduation. Just over 25 percent of currently enrolled veterans and non-veterans were in repayment because they had taken a break (stopped out) and then re-enrolled.6 Students enter repayment if they drop below half-time or fail to resume their studies within 6

4On average, veterans have a 5-year gap between high school graduation and when they begin or continue their higher education. 5The majority of veterans (58 percent) who enrolled for the first time in 2011-12 were 25 or older. 6The decision to re-enroll may have been influenced by the availability of loan payment deferments for individuals still in school.

6

months of stopping out. The implications of making student loan payments while still enrolled are daunting for both veterans and non-veterans. Such payments likely increase the use of deferments and may lead to delinquency and, potentially, default. They may also increase the pressure to work while enrolled or the number of hours worked; employment is a risk factor associated with noncompletion.

7

BACKGROUND This paper uses Beginning Postsecondary Students (BPS), a longitudinal survey conducted by the Department of Education’s National Center for Education Statistics (NCES) to examine factors contributing to and the impact of non-completion for “first-time” student veterans pursuing a postsecondary education—certificates, associate’s, or bachelor’s degrees.7 BPS includes student veterans who were first surveyed during the 2011-12 academic year and again in 2013-14. Because students surveyed by BPS for a second time in 2014 had only 3-years to earn a degree, we used the Department’s National Postsecondary Student Aid Study (NPSAS:16) to analyze time-to-degree from students’ first enrollment date, obtaining a longer-term perspective.8 For comparison, we include data on non-veteran independent students, a group comparable to veterans that we refer to as “non-veterans.”9 In general, both veterans and such non-veterans are financially independent, older, and are often married with dependent children. Because the focus is on veterans, we highlight the differences between veterans and non-veterans only when we believe they merit attention. See “Data Sources and Methodology” at the end of this paper for a more detailed discussion of BPS and our methodology. Use of GI Bill Benefits. Among veterans who used the GI Bill during the 2011-12 school year, about two-thirds received Post-9/11 GI Bill benefits; the remaining one-third continued to rely on older benefit

programs, primarily the Montgomery GI Bill.10 Under the Montgomery GI Bill, veterans receive a monthly benefit that can be used for tuition, housing, or other education related expenses. That payment would more than cover in-state tuition and fees at a community college or public 4-year institution but not tuition and fees at for-profit or nonprofit schools.11 In contrast, the Post-9/11 GI Bill pays tuition and fees directly to the school; during the 2011-

7Each cycle of the BPS follows a cohort of students who are enrolled in their first year of postsecondary education. For example, it would exclude a student who had transferred from or earned credits toward a certificate or degree at a different institution. The proportion of veterans who began school in 2011-12 and who completed a bachelor’s degrees by 2014 is less than 1 percent; most students pursuing bachelor’s degrees are still enrolled. 8BPS is a subset of first-time students who participated in the NPSAS survey in 2011-12. However, we used NPSAS: 16 to examine time-to-degree from student’s first enrollment date. 9The Education Department believes that non-veteran independent students are the most appropriate comparison group for veterans. 10From fiscal years 2009 through 2012, the proportion of eligible veterans using the Post-9/11 GI Bill to pursue a postsecondary education increased from 6 percent to 63 percent (GAO, 2013). When the benefit was first implemented, only 2 months remained in fiscal year 2009. Since 2012, the proportion of veterans using the Post-9/11 benefit has continued to increase at the expense of Montgomery GI Bill enrollment. 11In 2011-12, veterans using the Montgomery GI Bill received $1,473 a month for full-time attendance, totaling $13,257 over a 9-month academic year. According to the College Board, average tuition and fees were about

8

12 academic year, covered in-state tuition at public institutions and up to a cap of $17,500 at for-profit schools and nonprofit institutions.12 In addition, any gap between a school’s tuition and the Post-9/11 benefit can be reduced or eliminated if the school participates in the Yellow Ribbon Program.13 Under this voluntary program, the Department of Veterans Affairs (VA) matches the amount that the school provides to help close or eliminate any tuition gap.14 Unlike the Montgomery GI Bill, the Post-9/11 benefit offers a monthly housing stipend and pays $1,000 toward the cost of books.15 In addition to the GI Bill, veterans may qualify for Pell Grants and/or take out student loans. The Post-9/11 benefit does not reduce veterans’ eligibility for federal student loans. Both benefit programs pay for 36 months of postsecondary education.16 The benefit level is based on months-of-service. For example, veterans eligible for the full Post-9/11 benefit must have served for 3 years after September 10, 2001. If they served less than 3-years, the benefit is prorated in 10 percent increments from 40 percent up to 90 percent. Thus, tuition, housing, and book payments for a veteran who qualifies at the 40 percent level are reduced by 60 percent. According to the VA, about 30 percent of veterans who qualify for the Post-9/11 GI Bill receive less than 100 percent of the benefit. Two other factors can reduce the amount of the benefit—less than full-time attendance and enrollment in an exclusively-online program.17 For veterans attending school less than full-time (12 credit hours) housing expenses and the book stipend are prorated. Those pursuing 6 credits or less receive no housing stipend. Veterans attending classes exclusively online receive housing expenses equal to half the national average. Not All Enrolled Veterans Are Using the GI Bill. Not all veterans pursuing a postsecondary degree are using GI Bill benefits, a factor which may contribute to student loan debt. Overall, 37 percent of first-

$3,100 at a community college and about $7,770 at a public 4-year institution in 2011-12 (current dollars). For the 2018-19 academic year, the monthly payment is $1,994. 12The tuition cap is adjusted annually for inflation and is $23,672 for the 2018-2019 academic year. 13In Fiscal Year 2017, eighteen percent of schools that enroll 55 percent of GI Bill users participated in the Yellow Ribbon Program. However, only 4 percent of GI Bill recipients received Yellow Ribbon benefits. See “Veterans Initiative” and “Quick Facts” tabs of the “download data on all schools” link on the homepage of VA’s GI Bill Comparison Tool.

14Only veterans who qualify for 100 percent of the Post-9/11 benefit are eligible for the Yellow Ribbon program. Schools may limit the number of veterans who participate in the program or the amount they contribute. 15The monthly housing stipend is based on an enlisted rank of E-5 with dependents and is adjusted for the zip code of the school that the veteran attends. For example, the 2018 monthly allowance is $1,095 for Boise, Idaho, but $3,669 for New York City. According to NCES, the average cost of books during the 2012-13 academic year ranged from about $1,200 at public and nonprofit schools to about $1,700 at for-profit schools. The housing stipend can be used to cover any ongoing expenses such as food, day care, commuting expenses, etc. 16Thirty-six months is the equivalent of four, 9-month-long academic years—the time it should take to earn a 4-year bachelor’s degree. Veterans must attend full-time in order to graduate in 36 months.

17Full- or part-time attendance exhaust equivalent months of eligibility.

9

time student veterans were not using benefits in 2014 and 58 percent had cumulative student loan debt compared to 31 percent of benefit recipients.18 Reasons for not using benefits include:

• eligible but using state public school tuition waivers/scholarships intended for veterans,

• enrolled at an inexpensive community college and saving the GI Bill for a planned transfer to a 4-year institution,

• saving the benefit for a post-graduate degree,

• transferred the benefit or a portion of the benefit to a spouse or child, and

• ineligible because discharged from active duty prior to September 10, 2001, had too few months of qualifying service, or had a less-than-honorable discharge.

Exhausting the 36 months of benefits is another reason for not using the GI Bill. It is unlikely, however, that a veteran who enrolled for the first time in 2011-12 would have already exhausted 36 months of GI Bill benefits in 3 academic years (27 months). Because we can’t determine why veterans are not using GI Bill educational benefits, our analysis includes all veterans. Why Do Veterans Take on Student Loan Debt? While veterans not using GI Bill benefits may need to take on student loan debt, why do some benefit recipients borrow given the generosity of the Post-9/11 GI Bill? There are a number of other reasons, including:

• About 30 percent of veterans are only eligible for partial benefits because they served less than 3-years on active duty after September 10, 2001 and they are not eligible for the Yellow Ribbon Program to help cover any gap between benefits and the cost of college.

• Family circumstances, such as having a non-working spouse, children, or child care expenses may complicate some family’s financial situation.19

• Some veterans may take out low-interest federal student loans to pay down high-interest rate debt.

• Part-time attendance and enrolling in an exclusively-online program results in a reduction in or elimination of benefits for living and book expenses.

Where Are First-Time Veterans Enrolling? In 2011-12, most veterans (58 percent) enrolled in a public, postsecondary institutions—49 percent in 2-year or less institutions and 11 percent in 4-year institutions. Almost 40 percent of veterans enrolled in a for-profit school and 6 percent chose a nonprofit institution (see fig. 1). The enrollment pattern of non-veterans was similar, but a slightly higher percentage enrolled at public institutions (60 percent) and a smaller proportion enrolled in for-profit schools (34 percent).

18The corollary to those with debt is individuals who didn’t borrow. Thus, 42 percent of veterans not receiving benefits had no student loan debt; contributing factors are eligibility for Pell Grants, receipt of institutional aid, and the proportion of veterans enrolling in lower-cost community colleges. With respect to benefit recipients, about the same proportion of those who left or completed certificates or associate’s degrees were using benefits (67-68 percent). However, the proportion still enrolled and using benefits was somewhat lower—59 percent. By sector, a greater proportion of veterans were using benefits at for-profit schools than at public institutions in 2014. For example, 77 percent of veterans still enrolled were using benefits at for-profit schools compared to 53 percent at public-sector institutions. 19The living stipend is the same for single veterans and those who have dependents.

10

Figure 1: Distribution of Veterans Who Enrolled for the First Time in 2011-12 by Sector

Source: BPS:12/14. Calculations by authors.

58% 6% 37%

0% 20% 40% 60% 80% 100%

Total Veterans

Public Nonprofit For-profit

11

FINDINGS Data May Overstate Veteran Non-Completion Rates BPS data may overstate the impact of non-completion because many students take a temporary break (referred to as a “stop out”) or reenroll after a longer absence. The impact of such stop outs on completion is evident in the long-time-to-degree for both veterans and non-veterans, which suggests that some students persist and eventually earn a certificate or degree.

Veteran Non-Completion Rates. Overall, 20 percent of veterans left school without a degree after 3

years compared to 40 percent of non-veterans (see fig. 2). On average, both veterans and non-veterans left after fewer than 10 months. Twenty percent of veterans left both for-profit schools and public institutions without a degree. However, far fewer veterans were still enrolled at for-profit schools (27 percent) compared to their counterparts attending public institutions (68 percent). Figure 2 also shows that a greater proportion of veterans earned a certificate or degree by 2014 (28 percent) or were still enrolled (51 percent), compared to non-veterans (23 percent and 36 percent, respectively). Figure 2: Distribution of Attainment and Persistence as of 2014 Among First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12 by Sector

Source: BPS:12/14. Calculations by the authors. Note: “No degree, still enrolled” includes students who were enrolled for part of the year in 2013-14. The 0.5 percent of veterans who began school in 2011-12 and who completed a bachelor’s degrees by 2014 is shown in burgundy; most students pursuing bachelor’s degrees are still enrolled. We excluded the 2 percent of non-degree or other program students. Sample sizes were too small to report on nonprofit institutions.

Graduation rates for certificates and associate’s degrees were higher at for-profit schools than at public institutions. Several factors need to be kept in mind when comparing such graduation rates. Based on our analysis of College Scorecard data for 2015-16, 63 percent students earned certificates at 2-year for-profits while 72 percent of 2-year public graduates earned associate’s degrees; moreover, few 4-year

18%

5%

37%

16%

7%

30%

21%

10%

7%

15%

7%

6%

11%

5%

51%

68%

27%

36%

44%

22%

43%

20%

20%

20%

40%

43%

36%

29%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Total

Public

For-profit

Nonprofit

Total

Public

For-profit

Nonprofit

Vet

eran

sN

on-v

eter

ans

Attained certificate Attained associate's degree Attained bachelor's degree No degree, still enrolled No degree, left without return

Sample size too small

12

institutions even offer certificates. The higher proportion of certificate graduates at for-profits shown in the BPS data likely reflect the sector’s dominance in offering certificate programs. The disparity between certificate and associate’s degrees completed is also a function of the shorter time needed to earn a certificate and the fact that some for-profit schools require students in such programs to attend full-time.

Stop Outs. Twenty percent of veterans who enrolled in 2011-12 had stopped out at least once by 2014 (see fig. 3); fewer than 1 percent of veterans had stopped out two times or more. A slightly smaller proportion of non-veterans had stopped out—18 percent. Stop outs were more prevalent among those still enrolled (24 percent) or who had graduated (18 percent) than among those who left (13 percent). The lower rate of stop outs for those who left without a degree is likely due to the fact that, on average, veterans who left did so after fewer than 10 months. Stop outs were also more frequent among veterans pursuing certificates (18 percent) and associate’s degrees (26 percent) than among those enrolled in bachelor’s degree programs (9 percent). Lower stop-out rates among veterans seeking bachelor’s degrees suggest that more veterans should be encouraged to enroll in such programs.

Figure 3: First-Time Student Veterans Who Enrolled in 2011-12 and Who Experienced One or More Stop Outs by 2014

Source: BPS:12/14. Calculations by the authors.

Time-to-Degree. Given the propensity of veterans to stop out and then re-enroll, it is unclear if those who left will return at some point to earn a degree.20 Our examination of the time veterans take to earn a degree from the date they first enrolled demonstrates that veterans exhibit strong persistence, suggesting that some may be motivated to return and complete their education. We measured time-to-degree by examining data on the initial enrollment date of veterans and non-veterans who earned certificates or degrees in 2015-16. Because BPS completion rates focus on students who enrolled for the first time in 2011-12 and completed by 2014, we used data from the National Postsecondary Student Aid Study (NPSAS) to

20BPS:16 may allow researchers to determine if any veterans and non-veterans resumed their education and if additional students left without a degree. It is scheduled to be released in March 2019.

13%

24%

18%20%

0%

10%

20%

30%

40%

Left without return Still enrolled Attainedcertificate or degree

Total Veterans

13

analyze the time-to-degree for veterans and non-veterans who graduated in 2015-16.21 We also examined the number of institutions attended prior to earning a degree. According to NPSAS: 16, 41 percent of veterans, and more than 32 percent of non-veterans, who graduated in 2015-16 began their postsecondary education in 2005 or earlier (see fig. 4). These proportions varied based on the degree being pursued, suggesting the importance of degree aspirations. For example, 50 percent of veterans who earned bachelor’s degrees in 2015-16 first enrolled in in 2005 or earlier compared to 36 percent of associate’s degree and 33 percent of certificate program graduates.22 Figure 4: Distribution by First Enrollment Date of Student Veterans and Non-Veterans Who Completed Certificates or Degrees in 2015-16

Source: NPSAS: 16. Calculations by the authors Note: Years are calendar, not academic, years. 2015 includes any month from January to December 2015.

In addition to taking breaks, another factor increases time-to-degree—transferring to a different school. When students transfer some credits may not be accepted by a new institution. The need to retake the same classes is a particular problem for students who transfer from a for-profit to a public or nonprofit institution and can also be an issue when students transfer from a 2-year to a 4-year public institution

21NPSAS is a survey conducted every 4-years to collect data on how students finance their postsecondary education. NPSAS is more representative of all veterans enrolled in postsecondary education because some veterans may have earned some college credits before joining the military or used Defense Department educational benefits to take classes while on active duty. 22Until 2017, veterans had 15 years to use their Post-9/11 GI Bill benefits. The Forever GI Bill, enacted in August 2017, eliminated the time limit for individuals whose last discharge from active duty was on or after January 1, 2013. The time limit for veterans to use their educational benefits and the fact that benefits are limited to 36 months may help to explain why veterans who take a long time to graduate may not be using GI Bill benefits by the time they earn a degree.

4%

3%

17%

13%

20%

16%

28%

38%

32%

31%

V E T E R A N

N O N - V E T E R A N

1990 or earlier 1991 to 2000 2001 to 2005 2006 to 2010 2011-2015

14

(GAO, 2017).23 Among all first-time student veterans beginning in 2011-12, 27 percent had enrolled in more than one institution by June 2016, compared to 15 percent of non-veteran students.

Demographics of Veterans Who Left

Who Are First-Time Student Veterans and Who Left? Compared to all veterans who enrolled for the first time in 2011-12, the 20 percent of veterans who had left by 2014 were more likely to have been older, male, racial/ethnic minority, first in their family to go to college, without a high-school diploma, disabled, single or married with dependents, and working full-time (see table 1).24 As discussed later in this paper, several of these characteristics are associated with an increased risk of non-completion. Though not shown in table 1, non-veteran first-time students differed from veterans in several ways. More non-veterans were:

• Between the ages of 35 and 44 (14 percent vs. 9 percent);

• Women (62 percent vs. 13 percent);

• Single with children (38 percent vs. 20 percent); and

• Without a high-school diploma (27 percent vs. 13 percent). Table 1: Demographic Characteristics of First-Time Student Veterans Who Enrolled in 2011-12 and Veterans Who Left by 2014

All veterans Veterans who left

Distribution (percentage)

Distribution (percentage)

Age

<25 42 36

25-34 40 46

35-44 9 11

45+ 9 8

Gender

Male 87 95

Female 13 5

Race

White 61 52

Minority 39 48

1st generation 52 54

Family status Single (no dependents) 41 38

Married (no dependents) 15 9

Single (dependents) 20 27

Married (dependents) 23 26

No high school diplomaa 13 15

23Another factor that may contribute to veterans’ retaking courses is the higher likelihood that more selective institutions do not grant college credit for military training and occupations, regardless of institutional sector. For example, many nonprofit and public, flagship institutions do not grant college credits to veterans because they are more selective.

24When evaluating risk factors that may adversely affect persistence and attainment for nontraditional students, BPS classifies individuals with a GED as having “no high school diploma.”

15

Have a disabilityb 30 32

Employment status

Full-time 20 25

Part time 23 19

Not working 57 56

Source: BPS:12/14. Calculations by the authors. aWhen evaluating risk factors that may adversely affect persistence and attainment for nontraditional students, BPS classifies individuals with a GED as having “no high school diploma.” bBPS asked respondents four disability related questions: (1) are you deaf or do you have difficulty hearing; (2) are your blind or do you have a serious difficulty seeing even with glasses; (3) do you have a serious physical, mental, or emotional condition that causes serious difficulty concentrating, remembering, or making decisions; and (4) do you have serious difficulty walking or climbing stairs.

Age of Non-Completers by Sector. Figure 5 presents the results of a sector analysis of the ages of veterans and non-veterans who started their postsecondary education in 2011-12 and left by 2014. The data are based on the first institution in which these students enrolled. Veterans who left for-profit schools were older than those who left public institutions. For example, 73 percent of veterans age 25 or older left for-profit schools by 2014 compared to 62 percent of their peers at public sector institutions. Conversely, 39 percent of veterans who left public institutions by 2014 were 25 or younger compared to 27 percent who left for-profit schools. Non-veterans who left both for-profit and public institutions were much younger than their veteran counterparts—about 45 percent were under the age of 25. Figure 5: Age of First-Time Student Veterans and Non-Veterans in 2011-12 Who Left by 2014, Overall and by Sector

Source: BPS:12/14. Calculations by the authors.

Note: This figure shows the distribution of veterans and non-veterans who left postsecondary education by age based on the first institution in which they enrolled. Italicized data labels indicate sample sizes are small and should be interpreted with caution. Percentages along the axis report the distribution by sector within each group.

36%

39%

27%

0

46%

45%

45%

62%

46%

44%

53%

34%

33%

35%

27%

11%

11%

10%

13%

13%

14%

7%

8%

7%

10%

8%

9%

6%

5%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Total Veterans

Public (60%)

For-profit (37%)

Nonprofit (3%)

Total Non-veterans

Public (65%)

For-profit (31%)

Nonprofit (4%)

Under age 25 Age 25 to 34 Age 35 to 44 Age 45 or older

sample size too small

16

Factors Contributing to Non-Completion

Risk Factors. Researchers have identified seven factors, including several demographic characteristics, that can negatively affect persistence and attainment. These factors have been used by the Education Department to determine the risk of non-completion among nontraditional students (Molina and Morse, 2015). The seven factors are: delayed enrollment, no high school diploma, part-time enrollment, financially independent, have dependents, single parents, and full-time employment. We used five of these factors plus three others—first generation student, attendance at two or more institutions, and disability status—to assess the non-completion risk characteristics of student veterans and non-veterans in 2011-12.25 Compared to veterans, a greater proportion of non-veterans had 5 of the 8 risk factors—no high school diploma, enrolled part-time, had dependents, were single parents, and worked full-time while enrolled (see fig. 6). For example, 23 percent of non-veterans worked full-time compared to 20 percent of veterans. Research suggests that there is a complex relationship between working and student loan debt. A smaller proportion of those who work graduate with significant amounts of debt, and yet working too much while enrolled in a postsecondary program increases the likelihood of non-completion (Carnevale, et.al., 2015). With respect to the three other risk factors, the proportion of veterans and non-veterans sharing one risk factor was about the same but veterans were more at risk than non-veterans regarding the other two. Veterans and non-veterans were about equally likely to be first-generation students. The risk of non-completion for first-generation students is amplified because they share many of the risk characteristics associated with persistence and attainment challenges (Schauer, 2005). However, a much larger proportion of veterans than non-veterans had a disability (30 percent vs. 17 percent) or attended two or more institutions (27 percent vs. 15 percent). Disabled individuals are more likely to drop out than those without a disability (Velez, et.al., 2016).

25We did not use two factors: (1) financially independent—most GI Bill eligible veterans are no longer dependent on their parents; and (2) delayed enrollment—most veterans “delay postsecondary enrollment by more than 12 months” because they were on active duty for several years after high school and may not begin using their benefits immediately after separation from the military.

17

Figure 6: Distribution of Non-Completion Risk Factors Experienced by Student Veterans and Non-Veterans Who Enrolled for the First Time in 2011-12

Source: BPS:12/14. Calculations by the authors.

Note: BPS classifies individuals with a GED as having “no high school diploma” when defining risk factors.

Compared to those who graduated, veterans who left without a degree were much more likely to be associated with multiple risk factors (fig. 7). Thus, 21 percent of veterans who left school by 2014 without a degree had five or more non-completion risk factors compared to 4 percent of veterans who graduated. A greater proportion of non-veterans had 5 or more risk factors, both among those who left (29 percent) and those who graduated (17 percent). Figure 7: Distribution of Risk Factors by 2014 Completion Status for First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12

Source: BPS:12/14. Calculations by the authors.

Note: Italicized data labels indicate sample sizes are small and should be interpreted with caution.

Online Enrollment. Veterans are adult learners who often have families and jobs. As a result, online learning may be appealing because of the flexibility it offers to individuals who are juggling multiple responsibilities.

13%

24%

42%

17% 20%

30%

52%

27%27%

41%

57%

38%

23%17%

51%

15%

0%

20%

40%

60%

80%

100%

No high schooldiploma

Part-timeenrollment

Have dependents Single parent Working full-timewhile enrolled

Has a disability First generationcollege student

Attended 2 ormore institutions

Veterans Non-veterans

24%

22%

20%

17%

39%

27%

31%

22%

72%

62%

58%

55%

54%

52%

60%

55%

4%

17%

21%

29%

7%

21%

9%

23%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Veterans completing any degree

Non-veterans completing certificates or AA

Veterans who left (20%)

Non-veterans who left (40%)

Veterans still enrolled (51%)

Non-veterans still enrolled (36%)

All Veterans (8%)

All Nonveterans (92%)

1 or 2 3 or 4 5 or more

18

According to a 2017 Brookings report, however, online courses are especially difficult for those who are least prepared and, as a result, such students’ learning and persistence outcomes are worse than had they taken classes in-person (Bettinger and Loeb: 2017 and Bettinger, et.al., 2015).26 Because most veteran undergraduates use their GI Bill benefits after discharge from active duty, they may not have been exposed to an academic setting since leaving high school. Despite discipline and motivation, a lack of academic preparedness may require them to take remedial courses (Zoli, et.al., 2015). Financial stress may also contribute to non-completion among veterans enrolled in exclusively online programs. As noted earlier, such veterans don’t receive a full housing stipend; rather, their stipend is equal to half the national average. Although the proportion of veterans and non-veterans who left after taking classes exclusively online was similar (12 percent vs. 13 percent, respectively), the divergence between non-completion at for-profit and public institutions was stark (see fig. 8). About 20 percent of veterans attending for-profit schools in 2011-12 took all of their courses online compared to 7 percent of veterans attending public institutions.27 Among those who left by 2014, 36 percent of veterans had been enrolled exclusively online at for-profit schools compared to 19 percent of their peers at public institutions. Non-veterans who left were much less likely to be enrolled exclusively online at both for-profit schools (22 percent) and public institutions (9 percent) compared to their veteran counterparts. Figure 8: Distribution of Courses Taken Exclusively On-line by Sector Among All First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12 and Among Those Who Left by 2014

Source: BPS:12/14. Calculations by the authors. Note: Italicized data labels indicate sample sizes are small and should be interpreted with caution. Sample sizes were too small to report on nonprofit institutions.

26The researchers used data from a large for-profit school that offers the identical course online and in a classroom. 27Although for-profit schools pioneered online learning, distance education options are now widely available at public and some nonprofit institutions.

7%21%

12% 9%22%

13%19%

36%

12% 9%22%

13%

17%

8%13% 20%

6%14%

25%

11%

13% 20%6%

14%

76% 71% 76% 71% 72% 73%

56% 53%

76% 71% 72% 73%

0%

20%

40%

60%

80%

100%

Public For-profit Total Public For-profit Total Public For-profit Total Public For-profit Total

Veterans Non-veterans Veterans (20% who left) Non-veterans (40% who left)

All Some None

Among all students Among those who left by 2014

19

Implications of Non-Completion Research suggests that (1) individuals with student loan debt who leave school without a degree are more likely to face repayment challenges than those with debt who graduate (Akers and Chingos, 2014); and (2) student loan debt need not be excessive to cause financial stress in the short term for recent graduates and in the longer term for non-graduates (ACCT, 2015; Dynarski, 2014). Moreover, the likelihood of default is greater for those who do not graduate, compared to those who earn a degree because non-completers are less likely to obtain a job in their chosen career field and more likely to be unemployed (Gross, et.al., 2010; Hillman, 2015; Wei, 2013).

Student Loan Debt as of 2014. First, we examine cumulative student loan debt by degree level among veterans who borrowed and left by 2014 compared to those who also borrowed but graduated or were still enrolled.28 Then, we examine the distribution of that debt across the same three groups, including those who did not borrow. ➢ Cumulative Debt. As shown in fig. 9, a smaller proportion of veterans who left (28 percent) had

either federal or private student loan debt compared to veterans who graduated (39 percent) or were still enrolled (42 percent). Moreover, the average student loan debt of veterans who left was lower ($8,590) compared to their peers who graduated ($12,350) or were still enrolled ($9,900). In general, a larger proportion of non-veterans had student loan debt and higher average cumulative debt.

Figure 9: Federal and Private Cumulative Student Loan Debt as of 2014 by Attainment Status and Degree Program for First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12 and Who Borrowed

Source: BPS:12/14 and NPSAS:11/12. Calculations by the authors.

Note: Averages are rounded to nearest $10. Exact dollar amounts should be interpreted with caution. “Total” combines all students who left, regardless of degree program (certificate, associate’s or bachelor’s degree). Few students who enrolled in 2011-12 would have earned a bachelor’s degree within 3 years, that is, by 2014. For context, we report the cumulative student loan debt of students who earned bachelor’s degrees in 2011-12 using NPSAS:12. It is important to keep in mind, however, that

28We also examined cumulative debt for those receiving and not receiving GI Bill benefits. Thirty-one percent of veterans receiving benefits had student loan debt compared to 58 percent of their counterparts who were not using the GI Bill.

20

NPSAS and BPS data are not directly comparable because BPS focuses on first-time students, excluding those who had some college credits when they enrolled in 2011-12.

➢ Distribution of Cumulative Debt. Among veteran non-completers who borrowed, the largest

proportion (65 percent) had between $5,601-$19,000 in student loan debt (see fig. 10). Only 4 percent had borrowed more than $19,000 and 31 percent had less than $5,600 in student loan debt. In contrast, much higher proportions of veterans still enrolled or who graduated had cumulative loan debt totaling more than $19,000—24 percent and 20 percent, respectively. Almost three-quarters of veterans who left without a degree had no student loan debt, a much higher proportion than for those who graduated (60 percent) or those who were still enrolled (54 percent). A larger proportion of non-veterans who left had lower debt levels than veterans who did not complete.29

Figure 10: Distribution of Federal and Private Cumulative Student Loan Debt as of 2014 by Attainment Status for First Time Student Veteran and Non-Veteran Borrowers Who Enrolled in 2011-12

Source: BPS:12/14. Calculations by the authors. Note: The percentage of students who borrowed within each group is provided in parenthesis next to the degree completion status. Some certificate and associate’s degree graduates enrolled in bachelor’s degree programs and are shown here among those still seeking bachelor’s degrees. This figure excludes (1) the fewer than 3 percent of independent students enrolled in non-degree or unspecified programs; (2) the fewer than 1 percent of bachelor’s degree seekers who earned a degree within 3 years; and (3) veterans and non-veterans who did not borrow.

➢ Cumulative Debt by Sector. By sector, 36 percent of veterans who left for-profit schools without a degree had student loan debt compared to 24 percent of their peers as public institutions (see fig. 11); the average cumulative debt of veteran non-completers, however, was $1,600 less ($7,830 vs. $9,430, respectively). The overall proportion of non-veterans who borrowed (46 percent) was higher than that of veterans (28 percent), but their average student loan debt was lower.

29Overall, 59 percent of veterans and 42 percent of non-veterans had no student loan debt, irrespective of whether they left, were still enrolled, or had graduated.

33%

44%

9%

31%

11%

19%

14%

47%

43%

36%

49%

65%

64%

48%

33%

51%

24%

20%

42%

4%

25%

33%

53%

2%

0% 20% 40% 60% 80% 100%

Completed Certificate or Associate's Degree (40%)

Certificate or Associate's Degree program, still enrolled (42%)

BA Degree program, still enrolled (52%)

Left within 2 years (29%)

Completed Certificate or Associate's Degree (71%)

Certificate or Associate's Degree program, still enrolled (55%)

BA Degree program, still enrolled (77%)

Left within 2 years (47%)

Ve

tera

ns

No

n-v

ete

ran

s

$1 to $5,600 $5,601 to $19,000 $19,001 or more

21

Figure 11: Federal and Private Cumulative Student Loan Debt by Sector for First-Time Student Veteran and Non-Veteran Borrowers Who First Enrolled in 2011-12 and Who Left by 2014

Source: BPS:12/14. Calculations by the authors.

Note: The percentages in parentheses indicate the proportion of veterans and non-veterans who left. Due to small sample sizes, the average cumulative debt among veterans who left the public sector is an estimate. A similar estimate is not available for veterans who left the nonprofit sector. “Total,” however, includes both the public and private sectors. This figure combines all students who left, regardless of degree program (certificate, associate’s, or bachelor’s degree).

Impact of Non-Completers’ Student Loan Debt. Although average student loan debt for both veterans and non-veterans who left was lower than for graduates, average debt understates the repayment challenges some students face.30 As noted earlier, student loan debt need not be excessive to cause financial stress for non-graduates (Campbell and Hillman, 2015; Dynarski, 2014). In fact, 35 percent of students who default have less than $5,000 in student loan debt (Council of Economic Advisors, 2016). The data on income and expenses is limited because (1) they reflect metrics only 3 years after enrollment, (2) sample sizes were too small to look separately at certificate and associate’s degree graduates, and (3) it excluded data on those pursuing bachelor’s degrees, which generally require 4 years to complete. The 2015-16 iteration of BPS may address some of these limitations, such as providing data on those who earned bachelor’s degrees 5 years after enrolling. First, we compare the monthly income and expenses of those who left by 2014 and those who graduated. Then, we examine the distribution of student loan debt payments for veterans who left without a degree vs. all veterans in repayment. To address sample size issues, we combined certificate and associate’s degree students.31

30In general, non-completers carry far lower debt than graduating seniors (Hillman, 2015). 31The next iteration of this longitudinal survey—BPS: 12/14/16—is scheduled to be released in March 2019. It will examine students’ persistence and completion through academic year 2015-16.

24%

36%

28%29%

79%

69%

46%

$9

,43

0

$7

,83

0

*to

o s

mal

l

$8

,63

0

$6

,74

0

$7

,24

0

$1

1,3

80

$7

,29

0

0%

20%

40%

60%

80%

public

(20% of veterans,40% of non-veterans left)

for-profit

(20% of veterans,36% of non-veterans left)

nonprofit

(11% of veterans,29% of non-veterans left)

Total

(20% of veterans,40% of non-veterans left)

Veterans Non-veterans

Sam

ple

siz

e to

o s

mal

l.

22

➢ Monthly Income and Expenses. Veterans who left without a degree by 2014 earned only $70 less per month on average than veterans who earned degrees (see fig. 12). Analogously, non-veterans who left earned only $50 less than their peers who completed degrees. The similarity in average salaries between those who left and those who earned either certificates or associate’s degrees is not surprising. According to a 2011 study by the Center for Education and the Workforce at Georgetown University, the gap in monthly median earnings between those with some college, no degree and associate’s degree graduates was 11 percent ($333) while the gap between the later and bachelor’s degree completers was much higher—31 percent ($1,039) (Carnevalle, et.al., 2011). On average, the household expenses of veterans who left, including housing, car payments, day care, and student loan debt, totaled 44 percent of monthly income compared to 62 percent for veterans who graduated (see fig. 12). Student loan payments were 4.1 percent of monthly income for veterans who left compared to 5.3 percent for those who graduated.

Figure 12: Monthly Income and Expenses of First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12 and Were in Repayment by 2014

Source: BPS:12/14. Calculations by the authors.

Note: The distribution shown is among those in repayment and in the sample. Averages are rounded to the nearest $10. Exact dollar amounts should be interpreted with caution. To estimate 2014 daycare expenses, we used monthly reported costs from 2011-12 and used the CPI to inflate them to 2014 dollars. Due to sample size issues, data for child care for veterans are estimated using the overall average for the group. Because they had no income data, we excluded veterans and non-veterans who are still enrolled and in repayment.

➢ Distribution of Student Loan Payments for All Veterans in Repayment vs. Non-Completers.

Compared to all veterans (20 percent) and non-veterans (14 percent) in repayment by 2014, a much smaller proportion of those who left had monthly payments exceeding $200—2 percent of veterans and 5 percent of non-veterans (see fig. 13). Conversely, a larger proportion of veterans who left had student loan debt payments exceeding $100 a month (35 percent) than their non-veteran counterparts (27 percent). And, the proportion of veterans with monthly payments of $100 or less was lower (65 percent) compared to non-veterans (73 percent).

$120 $90 $130 $80

$640 $430

$500 $500

$310

$80 $150 $140

$326

$363$285 $354

$2,260

$1,900$1,950

$0

$500

$1,000

$1,500

$2,000

$2,500

Studentswho attained

any degreeby 2014

Studentswho left by

2014*

Studentswho attained

any degreeby 2014

Studentswho left by

2014 with2014 income

Monthly payment on student loans in 2014 (federal and private) Monthly rent or mortgage payment amount in 2014

Car payment amount in 2014 Monthly 2012 daycare costs (in 2014$, rounded)

Remaining available monthly income

Non-veterans Veterans

(19%) (63%) (20%) (21%)

$2,190

23

Figure 13: Distribution of Monthly Student Loan Payments Among All First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12 and Were in Repayment by 2014 and Among Those Who Left by 2014 and Were in Repayment

Source: BPS:12/14. Calculations by the authors. Note: Included are those in repayment with $0 actual monthly payments, which could be the result of participation in an income-based repayment plan. Italicized data labels indicate sample sizes are small and should be interpreted with caution.

Repayment Adversity. Although BPS:12/14 does not include data on default, it does collect data on repayment adversity, including deferment, forbearance, and delinquency.32 Eighty percent of veterans who left school by 2014 and were in repayment had requested a forbearance and 28 percent had deferred loan payments—higher proportions than non-veterans (see fig. 14).33 About 79 percent of veterans and 83 percent of non-veterans had experienced up to five delinquencies. Moreover, 19 percent of veterans and 12 percent of non-veterans had been delinquent 6 or more times, suggesting that they might be at risk of defaulting. In 2016, the Consumer Financial Protection Bureau reported that loan servicers were steering struggling borrowers into temporary forbearance and deferment options instead of providing information on and facilitating enrollment in income-based repayment plans (CFPB, 2016).

32The omission of default data is likely related to the relatively short period of time between first enrolling in 2011 and a student’s status by 2014. The next cycle of BPS covering academic year 2015-16, which is scheduled to be released in March 2019, should contain additional repayment data.

33Under the standard 10-year repayment plan, student borrowers facing adverse income shocks can apply for loan deferment (if they are unemployed) or a forbearance (if the amount exceeds 20 percent of their gross income) (Mueller and Yanellis, 2017). Although forbearance and deferment may help borrowers avert default, it comes at a price because their loan balances continue to grow. During deferment, borrowers are responsible for the accrued interest on unsubsidized loans only. In contrast, borrowers in forbearance are responsible for accrued interest on both subsidized and unsubsidized loan.

18%

31% 32%41%

28%

26%33%

32%

33%

28%

33%22%

20%14%

2% 5%

0%

20%

40%

60%

80%

100%

Veterans Non-veterans Veterans Non-veterans

Under $50 per month $51-100 $101-$200 Over $200

All students Left by 2014

24

Repayment challenges can begin prior to graduation. Just over 25 percent of currently enrolled veterans and non-veterans were in repayment because they had stopped out and then re-enrolled.34 Students who drop below half-time or fail to resume their studies within 6 months of stopping out enter repayment. The implications of having to make student loan payments while still enrolled are daunting for both veterans and non-veterans. Such payments likely increase the use of deferments and may lead to delinquency and, potentially, default. They may also increase the need to work while enrolled or the number of hours worked; employment is a risk factor associated with noncompletion. Sample sizes were too small to examine repayment adversity by institutional sector. It is important to note, however, that a Senate investigation found that many large for-profit schools had hired contractors to steer students who were at risk of default into deferment or forbearance in order to “manage” their institutions’ default rates (Senate HELP, 2012). Moreover, a recent report by the Government Accountability Office found that some schools hired consultants to manage their default rates in order to avoid losing access to federal student aid; these consultants encouraged borrowers with past-due payments to put their loans in forbearance (GAO, 18). Figure 14: Repayment Difficulties Among First-Time Student Veterans and Non-Veterans Who Enrolled in 2011-12, Left by 2014, and Were in Repayment by 2014

Source: BPS:12/14. Calculations by the authors. Note: This figure includes only those students who were repaying federal or private loans in 2014.

34The decision to re-enroll may have been influenced by the availability of loan payment deferments for individuals still in school.

28%

80%

98%

79%

19%23%

71%

95%

83%

12%

0%

20%

40%

60%

80%

100%

Ever had a deferment Ever had a forbearance Ever had a delinquency Had 1 to 5delinquencies

Had 6 or moredelinquencies

Veterans Non-veterans

25

CONCLUSIONS Based on our analysis of BPS:12/14, 20 percent of veterans who started school for the first time in 2011-12 had left without a degree by 2014. Because some students stop out and then return to school while additional students may decide to leave, non-completion results as of 2014 should be considered fluid. The 2015-16 cycle of BPS, scheduled to be released later in 2018, will update our understanding of non-completion among veterans who enrolled for the first time in 2011-12.

Veteran Resilience. One takeaway that is worth noting is the resilience of veterans whose persistence exceeds that of a comparable group—other independent students who are not veterans. Thus, non-completion was more prevalent among such non-veterans: 40 percent left school over the course of 3 years. Moreover, a much smaller proportion (36 percent) of non-veterans were still enrolled in 2014 compared to 51 percent of veterans.

Adequacy of Supports to Bolster Completion Rates. The multiple risk factors associated with veterans who leave without a degree offer clues about the underlying causes of non-completion and raise questions about the adequacy of institutional or federal support in place to foster completion (Molina and Morse, 2015). For example,

• Veterans who are single parents might benefit from campus-based day-care options.

• Veterans who enroll part-time may be doing so because of family obligations or the need to supplement their GI Bill benefits through employment. Such veterans might benefit from targeted institutional grants, participation in Education Department funded work-study programs, or help with day care.

• The high level of disability among veterans pursuing a postsecondary education—30 percent in BPS:12—suggests the need for an examination of the accessibility of VA and campus-based health services to GI Bill student veterans.

• Programs fostering persistence and attainment must be broad in scope. The experience of the CUNY Accelerated Study in Associate Programs (ASAP), suggest that addressing just one of the factors underlying non-completion may be insufficient, that is, institutional support must be more comprehensive.35

Finally, more attention should be paid to efforts that target the preparedness of veterans to pursue rigorous degree programs before they leave active duty and begin to use their GI Bill benefits. These efforts should not be limited to laudable, post-discharge programs such as service-to-school or the Posse Foundation’s veterans initiative that encourage enrollment at more selective schools. They should begin when high school graduates enlist and continue throughout their military service. For example, the Defense Department’s Transition Assistance Program currently allows servicemembers to take classes that help them earn a promotion to a higher rank. They could be reimagined as a pipeline that helps prepare high school graduates to realize the promise of their GI Bill benefits by focusing on post-service career goals and classes that will help them to avoid using their benefits on remedial courses.

35ASAP offers a robust package of resources and services designed to help students gain and maintain academic momentum, including, personalized advisement, tutoring, special class scheduling that allows students to attend classes in convenient time blocks to accommodate their work schedules, and immediately addressing any remedial needs.

26

REFERENCES Akers, Beth, and Matthew Chingos, 2014. “Is a Student Loan Crisis on the Horizon?” Washington, D.C.: Brown Center on Education Policy, Brookings Institution, June. www.brookings.edu/research/is-a-student-loan-crisis-on-the-horizon/

Bettinger, Eric, Lindsay Fox, Susana Loeb, and Eric Taylor, 2015. Changing Distributions: How Online College Classes Alter Student and Professor. Stanford Center for Education and Policy Analysis, Working Paper No. 15-10, October. https://files.eric.ed.gov/fulltext/ED580370.pdf Bettinger, Eric, and Susana Loeb, 2017. “Promises and Pitfalls of Online Education.” Washington, D.C.: Brookings Institution, Evidence Speaks Reports, Vol. 2, #15, June. https://www.brookings.edu/wp-content/uploads/2017/06/ccf_20170609_loeb_evidence_speaks1.pdf Campbell, Coleen, and Nicholas Hillman, 2015. “A Closer Look at the Trillion: Borrowing, Repayment, and Default at Iowa’s Community Colleges.” Washington, D.C.: Association of Community College Trustees, September. www.acct.org/files/Publications/2015/ACCT_Borrowing-Repayment-Iowa_CCs_09-28-2015.pdf Carnevale, Anthony P., Stephen J. Rose, and Ban Cheah, 2011. “The College Payoff: Education, Occupations, Lifetime Earnings.” Washington, D.C.: Georgetown University, Center on Education and the Workforce. https://cew-7632.kxcdn.com/wp-content/uploads/2014/11/collegepayoff-complete.pdf Carnevale, Anthony P., Nicole Smith, Michelle Melton, and Eric W. Price, 2015. “Learning While Earning: The New Normal.” Washington, D.C.: Georgetown University, Center on Education and the Workforce. https://cew.georgetown.edu/wp-content/uploads/Working-Learners-Report.pdf Consumer Financial Protection Bureau (CFPB), 2016. "Annual Report of the CFPB Student Loan Ombudsman: Transitioning from Default to an Income-based Repayment Plan." Washington, D.C.: October. https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/102016_cfpb_Transmittal_DFA_1035_Student_Loan_Ombudsman_Report.pdf Council of Economic Advisors, 2016. “Investing in Higher Education: Benefits, Challenges, and the State of Student Debt.” Washington, D.C.: Executive Office of the President of the United States, July. https://obamawhitehouse.archives.gov/sites/default/files/page/files/20160718_cea_student_debt.pdf Dynarski, Susan, 2014. “An Economist’s Perspective on Student Loans in the United States.” Washington, D.C.: Brookings Institute, Economic Studies at Brookings, September. www.brookings.edu/wp-content/uploads/2016/06/economist_perspective_student_loans_dynarski.pdf Government Accountability Office, 2013. “VA Education Benefits: VA Needs to Improve Program Management and Provide More Timely Information to Students.” Washington, D.C.: May. https://www.gao.gov/assets/660/654790.pdf Government Accountability Office, 2015. "Federal Student Loans: Education Could Do More to Help Ensure Borrowers Are Aware of Repayment and Forgiveness Options." Washington, DC: August. www.gao.gov/assets/680/672136.pdf

27

Government Accountability Office, 2018. “Federal Student Loans: Actions Needed to Improve Oversight of Schools’ Default Rates.” Washington, D.C.: April. https://www.gao.gov/products/GAO-18-163 Gross, Jacob, Osman Cekic, Don Hossler, and Nick Hillman, 2010. “What Matters in Student Loan Default: A Review of the Research Literature.” Journal of Student Financial Aid, Vol. 39, Issue 1, January. http://publications.nasfaa.org/cgi/viewcontent.cgi?article=1032&context=jsfa Hillman, Nicholas, 2015. “Borrowing and Repaying Student Loans.” Journal of Student Financial Aid, Vol. 45, Issue 3, November. http://publications.nasfaa.org/cgi/viewcontent.cgi?article=1588&context=jsfa

Molina, Dani, and Andrew Morse, 2015. “Military-Connected Undergraduates: Exploring Differences Between National Guard, Reserve, Active Duty, and Veterans in Higher Education.” Washington, D.C.: American Council on Education and NASPA (Student Affairs Administrators in Higher Education), November. http://www.acenet.edu/news-room/Documents/Military-Connected-Undergraduates.pdf Mueller, H., and Constantine Yannelis, 2017. “Students in Distress: Labor Market Shocks, Student Loan Default, and Federal Insurance Programs.” Cambridge, MA: National Bureau of Economic Research, March 2017. http://papers.nber.org/tmp/57937-w23284.pdf Schauer, Ilya, 2005. “Issues Facing First Generation College Students.” Manhattan, Kansas: National Academic Advising Association (NACADA), Kansas State University, February. http://www.nacada.ksu.edu/Resources/Academic-Advising-Today/View-Articles/Issues-Facing-First-Generation-College-Students.aspx U.S. Senate Health, Education, Labor and Pensions Committee, 2012. “For-Profit Education: The Failure to Safeguard the Federal Investment and Ensure Student Success.” Washington, D.C.: United States Senate, Majority Committee Staff Report and Accompanying Minority Committee Staff Views, July. https://static1.squarespace.com/static/556718b2e4b02e470eb1b186/t/56112384e4b0521c793b0c4d/1443963780937/harkins-full-report.pdf Velez, Erin, Alexandria Radfod, Alexander Bentz, Stephen Lew, and Nicole Ifill (2016). “First-Time Postsecondary Students in 2011-12: 3-Year Retention and Attainment at First Institution.” Web Tables. NCES 2016-137. Washington, D.C.: National Center for Education Statistics (NCES), September. https://nces.ed.gov/pubs2016/2016137.pdf Wei, Christina Chang, and Laura Horn, 2013. “Federal Student Loan Debt Burden of Noncompleters.” Stats in Brief. NCES 2013-155. Washington, D.C.; National Center for Education Statistics (NCES), April. https://nces.ed.gov/pubs2013/2013155.pdf Zoli, Corri, Rosalinda Maury, and Daniel Fay, 2015. “Missing Perspectives: Servicemembers’ Transition from Service to Civilian Life.” Syracuse, N.Y.: Institute for Veterans and Military Families, Syracuse University, November. http://surface.syr.edu/cgi/viewcontent.cgi?article=1006&context=ivmf

28

DATA SOURCES AND METHODOLOGY

Data Sources. This paper uses Beginning Postsecondary Students Longitudinal Study (BPS:12/14) conducted by the Department of Education’s National Center for Education Statistics, to examine factors contributing to and the impact of non-completion for first-time student veterans pursuing a postsecondary education. Each cycle of the BPS follows a cohort of students, which includes veterans, who are enrolling in postsecondary education for the first time. The study collects data on student persistence in, and completion of, postsecondary education programs; their transition to employment; demographic characteristics; and changes over time in their goals, marital status, income, and both household and student loan debt, among other indicators. BPS includes students who began their postsecondary education in academic year 2011-12, with follow-up surveys in 2014 and 2017. NCES expects to release data from the most recent survey cycle (2016-2017) in March 2019. A total of 35,540 respondents were identified in the NPSAS:12 survey as potential first-time beginning postsecondary students and became the sample for the BPS:12/14 longitudinal study. In 2014, about 24,700 students completed the interview, resulting in a nationally-weighted interview response rate of 68 percent. The BPS:12/14 study sample represents the approximately 4-million undergraduates who were first-time postsecondary students in 2011–12. Respondents are asked if they are veterans, allowing researchers to differentiate between veterans and other GI Bill recipients—eligible family members, survivors, and service members using their benefits

while on active duty.36 BPS also uses school administrative data to determine veteran status. In addition to the student record collection and interview, NCES uses its student loan database, the National Student Loan Database System (NSLDS), for the types and amounts of federal loans and Pell Grants received by survey participants. Finally, the National Student Clearinghouse (NSC) provides information on postsecondary enrollment, degree, and certificate records on behalf of participating postsecondary institutions. It is important to emphasize that BPS looks at students who enrolled for the first-time in the 2011-2012 academic year. A different quadrennial NCES survey, the National Postsecondary Student Aid Study, offers a more comprehensive examination of students and student veterans because it also includes individuals who are not first-time students.

36The Free Application for Federal Student Aid (FAFSA) introduced a skip pattern in 2008 or 2009 and servicemembers, the Guard and Reserves, and veterans are no longer asked to self-identify if they are older, married, or have children—criteria that describe the majority of student veterans.

29

Methodology. This paper focuses on veterans who are undergraduates and compares their characteristics to other independent students who are not veterans.37 Dependent students are excluded because they are younger and considered financially dependent on their parents. In contrast, both veterans and non-veteran independent students are financially independent, older, and are often married with dependent children. By definition, BPS focuses on undergraduates, excluding students who are pursuing postgraduate degrees.

NCES provides two options in BPS for measuring persistence and attainment over the entire 3-year period—PROUT and PRAT.

• PROUT3 measures “as of the 2013-14 academic year,” reporting separately on students who were enrolled at some time during 2013-14, but not during the last few months.

• PRAT3Y, in contrast, measures “as of June 2014,” the end of the academic year, counting such students as having left.

The differences in the two variables is evident in the 7 percent difference in non-completion rates among veterans. PRAT3Y has a non-completion rate of 27 percent. For this paper, we used PROUT3 and report students not enrolled in the last few months of the academic year as “still enrolled during 2013-14.” This decision, which results in a 20 percent non-completion rate, is based on the fact that 20 percent of veterans stopped out between their enrollment in 2011 and 2014, taking a break from school of at least five consecutive months. Because they may have simply stopped out, we believe it is premature to conclude that students who were not enrolled for the last few months leading up to June 2014 had left.

We used Italicized data labels in some figures to alert readers to small sample sizes that should be interpreted with caution. In general, sample sizes were too small for nonprofit institutions. NCES’s own publications generally avoid sampling issues because they focus on all students surveyed—both dependent and independent students.

The estimates cited in this report were derived using PowerStats, the NCES web-based analytical software.

37NCES believes that non-veteran independent students are the most appropriate comparison group for student veterans.